Responsible Use of Generative AI in Lifestyle Medicine Education in Viet Nam: Accuracy, Transparency, and Human Oversight

Responsible Use of Generative AI in Lifestyle Medicine Education in Viet Nam: Accuracy, Transparency, and Human Oversight

Primary category: Lifestyle Medicine in Viet Nam

Secondary category: Research & Implementation

Article type: VLMA Perspective

Last reviewed: October 29, 2026

A healthcare educator asks a generative artificial intelligence tool to prepare a teaching session on nutrition and type 2 diabetes. Within seconds, the tool produces learning objectives, a presentation outline, practical examples, and a list of scientific references.

The material looks professional. The explanations are clear, and the references appear credible. With a few edits, it might seem ready to use.

But what if one of those references does not exist? What if a recommendation comes from an outdated guideline? What if the tool has presented an association identified in an observational study as proof that a particular intervention causes better health outcomes?

These errors may not be obvious to someone reading the finished presentation. Fluent language and professional formatting can make inaccurate information appear more reliable than it is.

Generative artificial intelligence is changing how educational content can be prepared, translated, illustrated, and shared. For healthcare professionals, it offers opportunities to work more efficiently and make scientific knowledge more accessible. It also raises important questions about accuracy, transparency, and responsibility.

The challenge is not simply whether artificial intelligence can help prepare a lecture or draft an article. It is whether the resulting material remains scientifically sound, appropriately sourced, and subject to meaningful human oversight.

For the Vietnam Lifestyle Medicine Alliance (VLMA), this question is particularly relevant.

VLMA is a physician-led professional platform operated by Vietnam Lifestyle Medicine Alliance Company Limited. Its direction is to support evidence-based Lifestyle Medicine in Viet Nam through professional education, capability development, scientific knowledge exchange, and collaboration.

Generative artificial intelligence may offer useful support for some of these activities. However, efficiency should not come at the expense of scientific integrity, professional responsibility, or public trust.

This article examines the responsible use of generative artificial intelligence in Lifestyle Medicine education, with particular attention to scientific accuracy, transparency, confidentiality, intellectual property, and the legal framework in Viet Nam.

Its focus is educational content development and publication rather than artificial intelligence systems intended for clinical diagnosis or individualized treatment decisions.

A Useful Tool Is Not an Authoritative Source

Generative artificial intelligence systems can produce text, images, presentations, translations, summaries, and other forms of content. Some tools can also retrieve information, process uploaded documents, or assist with tasks involving several sources.

These capabilities can be valuable in professional education.

An educator may use a tool to organize learning objectives, prepare discussion questions, or explore different ways to explain a complex concept. A researcher may use it to identify search terms or organize notes from previously collected publications. A professional organization may use it to prepare an initial translation or improve the readability of an educational resource.

For organizations with limited editorial or administrative capacity, such assistance can reduce some of the practical work involved in preparing content.

However, the usefulness of a tool does not establish the reliability of everything it produces.

Generative artificial intelligence can create statements that are incorrect, incomplete, misleading, or unsupported by the references provided. It may generate a convincing article title, journal name, and digital object identifier for a publication that does not exist.

It may also describe a genuine study inaccurately, combine findings from different sources without preserving important distinctions, or omit limitations concerning the population studied and the outcomes measured.

These problems can occur even when the output appears coherent and confident.

The risk is particularly relevant to medical education because inaccurate information may be repeated in teaching materials, professional presentations, online articles, or subsequent publications. Once a claim appears in several professional-looking resources, readers may assume that it has an established scientific basis.

Artificial intelligence should therefore not be treated as an independent scientific authority.

A generated explanation is not equivalent to an original research article, systematic review, clinical guideline, or verified professional standard. Even when a tool supplies citations, those references need to be checked against the actual sources.

The appropriate role of generative artificial intelligence is to assist work that remains grounded in verifiable evidence and accountable professional judgment.

Why Scientific Verification Matters in Lifestyle Medicine

Lifestyle Medicine draws on evidence concerning nutrition, physical activity, sleep, stress management, avoidance of risky substances, social connection, behavior change, and the prevention and management of chronic disease.

Educational resources in this field often bring together findings from clinical trials, observational studies, systematic reviews, professional guidance, and implementation research.

These sources do not necessarily answer the same questions or provide the same level of certainty.

Evidence that regular physical activity benefits health, for example, does not establish that every exercise program is effective or appropriate for every clinical condition.

Evidence supporting a dietary pattern does not validate every supplement, food product, or commercial program associated with that pattern.

Similarly, an intervention that improves an intermediate clinical measure may not have demonstrated an equivalent effect on long-term health outcomes.

These distinctions matter when healthcare professionals are learning how to apply Lifestyle Medicine responsibly.

A generative artificial intelligence tool may produce a clear explanation while overlooking the qualifications that make it scientifically accurate.

It may exaggerate the magnitude of an intervention's benefit, fail to distinguish association from causation, or present preliminary findings as established clinical recommendations.

It may also generate advice that appears reasonable in general but is inappropriate for certain patient populations.

Educational material about physical activity, for instance, should not overlook relevant precautions or contraindications. Nutrition education must recognize that some patients require individualized dietary management because of chronic kidney disease, pregnancy, nutritional deficiencies, or other clinical conditions.

Accuracy therefore involves more than checking whether individual statements are factually correct.

It requires understanding the quality of the evidence, the relevance of the source, the strength of the conclusion, and the circumstances in which the information may be applied.

For clinical recommendations, this generally means consulting the original guideline or authoritative professional guidance, including its publication date, intended population, scope, and important qualifications.

For research findings, it means checking the actual publication, study design, participants, interventions, outcomes, and limitations before using the results to support a claim.

Particular care is needed when educational content is described as presenting the latest evidence.

A generative artificial intelligence system may not have access to a recently published guideline. Even a system capable of retrieving information may fail to distinguish updated recommendations from older versions.

A retrieved source is not necessarily current, authoritative, or applicable to the intended setting.

For educational resources intended for healthcare professionals in Viet Nam, scientific appraisal also needs to consider local relevance.

An international recommendation may require interpretation in light of available healthcare services, professional scopes of practice, patient circumstances, and Vietnamese regulations.

Artificial intelligence may help identify questions worth investigating, but those questions must be evaluated by people with appropriate expertise.

Scientific verification remains a human responsibility.

Translation and Educational Adaptation Require Professional Judgment

Language support is one of the potentially useful applications of generative artificial intelligence in international professional education.

Many important Lifestyle Medicine resources are published in English. Accurate Vietnamese explanations can help make this knowledge more accessible to healthcare professionals working in Viet Nam.

Artificial intelligence can assist with initial translation, terminology comparisons, plain-language summaries, and the adaptation of educational material for different audiences.

However, a fluent translation is not necessarily a clinically accurate translation.

A small change in wording may alter the strength of a recommendation, the description of a risk, or the meaning of an important clinical qualification.

The distinction between an intervention that may be beneficial and one that is recommended can matter considerably.

Similarly, terms such as remission, reversal, prevention, treatment, contraindication, and referral should not be treated as interchangeable.

A translation may preserve the words of a guideline while losing the conditions under which its recommendations apply.

Educational material should therefore be reviewed for both linguistic accuracy and clinical meaning.

Where terminology differs between international publications and Vietnamese professional practice, those differences may need to be explained rather than concealed.

The same principle applies when simplifying material for learners.

An introductory resource may reasonably omit some technical details, but it should not remove important limitations or create an impression of certainty that the evidence does not support.

A resource intended for healthcare professionals must also distinguish general scientific education from activities requiring clinical assessment, professional authorization, or appropriate supervision.

Generative artificial intelligence can help make knowledge easier to communicate. Qualified educators and relevant subject-matter experts must still determine whether the resulting material remains faithful to the evidence.

Authorship, Transparency, and Responsibility for Published Content

International scientific publishing guidance increasingly addresses the use of artificial intelligence.

In January 2026, the International Committee of Medical Journal Editors updated its Recommendations for the Conduct, Reporting, Editing, and Publication of Scholarly Work in Medical Journals.

Section V addresses the use of artificial intelligence in publishing, including the responsibilities of authors, reviewers, editors, and publishers.

An important principle is that artificial intelligence tools should not be listed as authors because they cannot accept responsibility for the accuracy, integrity, and originality of published work.

Human authors remain responsible for material prepared with artificial intelligence assistance.

The recommendations also emphasize disclosure of artificial intelligence use, appropriate attribution, source verification, protection of confidential manuscripts, and prevention of plagiarism.

For scholarly medical publications, contributors are expected to disclose how artificial intelligence tools were used, including the relevant tool and purpose.

These considerations matter because artificial intelligence may contribute to content development in different ways.

Using a tool to correct spelling or grammar is different from relying on it to interpret clinical studies, generate scientific explanations, select references, or draft substantial sections of an educational resource.

The more substantially a tool contributes to scientific content or interpretation, the more important it becomes to examine accuracy, originality, disclosure, and responsibility.

The International Committee of Medical Journal Editors' recommendations are intended primarily for scholarly medical publishing. They do not automatically impose identical requirements on every educational website or professional organization.

Nevertheless, their underlying principles are relevant to VLMA.

An article published on a professional website may not be a peer-reviewed journal publication, but it can still influence healthcare professionals' understanding of scientific evidence.

Readers should be able to trust that the material has been prepared responsibly, that its references are genuine, and that accountable people stand behind the published content.

Transparency about artificial intelligence use should therefore be meaningful rather than merely symbolic.

A general statement that artificial intelligence was used may tell readers little about what the tool actually contributed.

Where artificial intelligence has materially assisted with substantive content, an organization should consider explaining its role and the human review performed, taking account of applicable legal and editorial requirements.

Such a statement should describe what actually happened. It should not imply that scientific review was performed when it was not, or that an article was independently peer-reviewed merely because an artificial intelligence system examined the text.

Artificial intelligence assistance does not remove the publisher's responsibility for what is published.

Synthetic Images and the Risk of Misrepresenting Reality

Visual materials can make medical education more accessible. Illustrations, diagrams, and educational graphics can help explain concepts that are difficult to communicate through text alone.

Generative artificial intelligence has made it easier to create such materials, including realistic images of healthcare professionals, patients, clinical environments, and educational activities.

These images can be appropriate when used as illustrations.

Problems arise when a synthetic image appears to document something that did not actually happen.

An artificial intelligence-generated image of a healthcare workshop should not be presented as evidence that an organization held that workshop.

A realistic illustration showing physicians and patients in a hospital should not suggest that VLMA operates a clinical service or has a formal relationship with an identifiable institution when no such activity or relationship exists.

The same concern applies to synthetic certificates, professional seals, accreditation symbols, testimonials, and depictions of identifiable individuals.

Visual communication should not create factual claims that the accompanying text does not support.

Where an artificial intelligence-generated image could reasonably be mistaken for a photograph of an actual event or clinical activity, its illustrative nature should be made clear.

For example, an accompanying caption could explain that the image is an artificial intelligence-generated illustration and does not depict an actual VLMA event or patient encounter.

The precise legal requirements depend on the content and circumstances.

Viet Nam's Law on Artificial Intelligence No. 134/2025/QH15, effective March 1, 2026, establishes transparency responsibilities concerning certain artificial intelligence systems and generated content.

Article 11 distinguishes several obligations.

Providers have responsibilities concerning users' ability to recognize when they are interacting directly with artificial intelligence systems and the machine-readable marking of certain generated audio, images, and videos.

Deployers have responsibilities to provide appropriate public notification when artificial intelligence-generated or modified content could mislead people about the authenticity of events or individuals.

Visible labeling requirements also apply to certain synthetic audio, images, and videos that simulate real individuals or recreate actual events.

Decree No. 142/2026/ND-CP, effective May 1, 2026, provides further provisions. Article 17 addresses technical marking, while Article 18 addresses notification and visible labeling.

The decree also identifies circumstances in which visible labeling is not required, including certain technical edits and text processing activities involving spelling correction, grammar, summarization, paraphrasing, or translation that do not materially distort the original content.

These distinctions matter.

It would be inaccurate to claim that every use of an artificial intelligence writing assistant automatically requires the same public label.

Equally, an exception from visible labeling should not be interpreted as a general exemption from all applicable transparency, copyright, or professional responsibilities.

The relevant obligations depend on the activity, the type of content, the legal role of the party involved, the risk of misleading the public, and the applicable statutory provisions.

For VLMA, a sensible principle is to distinguish clearly between illustration and documentation.

Scientific credibility should extend to images as well as written content.

Protecting Patient Information and Confidential Materials

Generative artificial intelligence tools may require users to enter information, upload documents, or provide access to files and other data.

This creates particular concerns when educational material involves clinical records, patient experiences, unpublished research, or internal organizational documents.

A healthcare professional preparing a teaching session may wish to summarize a real patient case. An educator may want assistance analyzing participant feedback. An editor may consider uploading an unpublished manuscript for language editing.

These tasks can involve information that should not be disclosed to an external artificial intelligence service without appropriate assessment and authorization.

Removing a patient's name is not always enough to eliminate identification risks.

Clinical characteristics, dates, unusual circumstances, photographs, and combinations of seemingly ordinary details may still allow someone to be identified.

Unpublished documents may also contain confidential institutional information, intellectual property, or material protected by contractual obligations.

Using external artificial intelligence services can introduce additional questions about data access, storage, retention, security, and international transfers.

In Viet Nam, the Law on Personal Data Protection No. 91/2025/QH15 and Decree No. 356/2025/ND-CP, both effective January 1, 2026, establish requirements relevant to the processing and protection of personal data.

The applicable obligations depend on the type of information, the processing purpose, the legal basis, the organizations involved, and other circumstances specified by law.

Consent may be necessary in particular situations, but it should not be assumed to be the only possible legal basis in every case.

Conversely, the fact that a healthcare organization already possesses patient information does not automatically authorize its disclosure to an external artificial intelligence provider for an unrelated purpose.

Professional confidentiality, institutional permissions, contractual restrictions, security safeguards, and other applicable obligations must also be considered.

A prudent approach is to avoid entering identifiable patient information, confidential clinical records, unpublished manuscripts, or restricted institutional documents into external tools unless the proposed use has been appropriately assessed and authorized.

Where artificial intelligence assistance is needed, organizations should consider whether the task can be completed using publicly available information, properly prepared fictional examples, or appropriately de-identified material.

Even then, confidentiality and re-identification risks may remain.

The fact that a tool is easily accessible does not make it appropriate for every kind of information.

Copyright and the Use of Artificial Intelligence-Generated Material

Generative artificial intelligence can produce text and images that appear original. However, the fact that material was generated by a tool does not automatically resolve questions of copyright, ownership, attribution, or permission.

A system may generate language resembling existing publications, reproduce recognizable elements of protected visual works, or summarize copyrighted material without identifying its source.

These concerns become particularly relevant when educational content is intended for public distribution.

Healthcare educators should distinguish between discussing scientific findings and reproducing protected expression.

Research findings can be critically interpreted and accurately summarized, but copying substantial passages, tables, figures, or proprietary educational materials may raise separate copyright concerns.

Translation and adaptation can also involve rights associated with the original work.

Artificial intelligence assistance does not create a general exemption from these responsibilities.

In Viet Nam, the Law on Intellectual Property No. 50/2005/QH11, as amended, provides the relevant legal framework.

An important recent development is Law No. 131/2025/QH15, adopted on December 10, 2025, and effective April 1, 2026.

The amendment addresses issues arising from artificial intelligence, including the establishment of intellectual property rights for subject matter created with the assistance of artificial intelligence systems.

It also provides for further government regulation of relevant circumstances.

Decree No. 134/2026/ND-CP, issued on April 6, 2026, and effective April 9, 2026, further amended provisions governing copyright and related rights under Decree No. 17/2023/ND-CP.

Of particular relevance is the introduction of Article 5a, which addresses copyright and related rights where artificial intelligence systems are used in the creative process.

The regulatory approach recognizes the importance of identifiable human creative contribution, including creative choices and control over the resulting work.

This does not mean that every artificial intelligence-generated output automatically receives the same protection as a human-created work.

Nor does it mean that generated material is automatically free for unrestricted commercial use.

Questions concerning protection, ownership, permissions, and possible third-party rights depend on the circumstances and applicable legal provisions.

Scientific integrity and copyright compliance are related but distinct responsibilities.

An educational resource may be scientifically accurate while still raising copyright concerns. Permission to reproduce an image or document, on the other hand, does not establish that its content is scientifically reliable.

Both issues deserve attention before publication.

Applying Viet Nam's Artificial Intelligence Framework Responsibly

Viet Nam's legal framework for artificial intelligence has developed substantially.

The Law on Artificial Intelligence No. 134/2025/QH15 was adopted on December 10, 2025, and entered into force on March 1, 2026.

The law addresses the research, development, provision, deployment, and use of artificial intelligence systems, together with the rights and responsibilities of relevant organizations and individuals.

It establishes a human-centered approach and a framework for managing artificial intelligence systems according to risk.

Decree No. 142/2026/ND-CP, issued on April 30, 2026, and effective from May 1, 2026, provides further implementing provisions.

These developments are relevant to organizations using generative artificial intelligence for educational activities.

However, the legal obligations applicable to a particular organization cannot be determined solely by identifying the technology as generative artificial intelligence.

The organization's role, the functions of the system, the purpose of use, the characteristics of the content, and the potential consequences must all be considered.

An organization using a general-purpose tool to help prepare a public educational article is not necessarily subject to the same obligations as an entity developing or deploying an artificial intelligence system intended for clinical diagnosis.

The law distinguishes the roles and responsibilities of developers, providers, deployers, and other relevant parties.

Using a third-party tool does not, by itself, establish that an organization is the developer or provider of that system. However, its actual activities may still create responsibilities under applicable law.

Educational activities can also involve other legal requirements.

Personal data protection obligations may arise when information about identifiable individuals is processed. Intellectual property requirements may apply when source material or generated content is reproduced or distributed.

Advertising and consumer protection provisions may be relevant where published content promotes commercial products or services.

Healthcare legislation and professional scopes of practice remain important where material could be interpreted as recommending clinical interventions or providing individualized medical care.

Responsible use therefore requires attention to the actual activity rather than reliance on a general claim that a tool is safe, compliant, or suitable for professional use.

For VLMA, these legal developments provide an important context for professional education. They do not automatically make VLMA a developer, provider, or certifier of artificial intelligence systems merely because third-party tools may be used to support selected editorial tasks.

At the same time, using an available technology does not establish that all applicable legal, professional, or contractual requirements have been satisfied.

Human Oversight Must Be More Than a Final Approval

Human oversight is often presented as the main safeguard against inaccurate artificial intelligence-generated content.

However, simply having someone approve the final document does not mean that its scientific content has been reviewed adequately.

A superficial review may detect grammatical errors while missing invented references, misleading interpretations, or inappropriate clinical recommendations.

Meaningful oversight requires an understanding of the intended audience, the subject matter, and the limitations of the available evidence.

For a Lifestyle Medicine educational resource, the responsible reviewer should be able to assess whether important claims are supported by appropriate sources, whether clinical qualifications have been preserved, and whether the conclusions are consistent with the evidence.

The review should also consider whether the material could be mistaken for individualized medical advice, whether commercial interests have influenced its presentation, and whether confidential information has been handled appropriately.

The level of review should reflect the potential consequences of error.

An introductory explanation of a general scientific concept may not require the same specialist review as a resource discussing medication adjustment, contraindications, or care for patients with multiple chronic diseases.

Where specialized clinical expertise is necessary, appropriately qualified professionals should be involved.

Human oversight also means having the authority to reject generated material.

If an artificial intelligence output is unreliable, misleading, or insufficiently supported, discarding it may be more appropriate than continuing to revise it.

This is particularly relevant when a tool produces a long document that appears polished but contains numerous unsupported assertions.

The objective should not be to preserve as much generated text as possible. It should be to produce an educational resource that meets the required scientific and professional standards.

International guidance supports this approach.

The World Health Organization's 2024 publication, Ethics and Governance of Artificial Intelligence for Health: Guidance on Large Multi-Modal Models, addresses risks involving inaccurate information, bias, automation bias, privacy, and accountability.

The United Nations Educational, Scientific and Cultural Organization's Guidance for Generative AI in Education and Research, published in 2023, provides a complementary perspective on human-centered use, educational capacity, privacy, and critical evaluation.

These documents are international guidance resources rather than substitutes for Vietnamese legislation.

Their relevance lies in recognizing that responsible technology use depends not only on what a system can produce, but also on how people assess, supervise, and communicate its outputs.

What This Means for VLMA

For VLMA, generative artificial intelligence presents both an opportunity and a responsibility.

As a developing professional platform, VLMA aims to make evidence-based Lifestyle Medicine knowledge more accessible to healthcare professionals in Viet Nam.

Tools that assist with translation, content organization, educational design, and communication may help support that work.

However, efficiency should not become the main measure of success.

An educational resource that is carefully prepared, appropriately sourced, and professionally reviewed may provide greater value than a large quantity of content produced quickly without adequate verification.

For VLMA, the priority should remain the quality, relevance, and trustworthiness of information.

As its activities develop, one possible direction would be to distinguish between editorial tasks that can reasonably be supported by artificial intelligence and those requiring substantive professional judgment.

It may also be appropriate to develop proportionate practices concerning reference verification, human editorial responsibility, disclosure of material artificial intelligence assistance, protection of confidential information, and correction of published errors.

Such practices would need to reflect the organization's actual activities, available expertise, and applicable legal requirements.

They should not be presented as a comprehensive artificial intelligence governance system or formally adopted institutional policy unless VLMA has genuinely established and implemented them.

VLMA does not operate a clinical artificial intelligence service, approve artificial intelligence products, or provide regulatory certification.

Its intended contribution in this area is professional education and responsible knowledge exchange.

This includes helping healthcare professionals understand both the opportunities and limitations of generative artificial intelligence when preparing, evaluating, or using Lifestyle Medicine educational resources.

The aim is not to discourage appropriate technological assistance. It is to ensure that new tools strengthen professional work without weakening the standards on which credible medical education depends.

Using Technology Without Losing Professional Judgment

Generative artificial intelligence can help people work with information, but it does not remove the need to understand that information.

It may organize a lesson, suggest clearer language, or assist with translation. It may also produce inaccurate references, overlook important limitations, or turn uncertain evidence into confident statements.

The difference between useful assistance and misleading content often depends on how carefully the output is evaluated.

For healthcare professionals, this responsibility matters because educational material may influence clinical understanding and professional decisions long after it has been published.

Accuracy requires checking the evidence rather than trusting a fluent explanation. Transparency requires communicating the role of artificial intelligence when appropriate and legally required. Human oversight requires people who can question, correct, and reject generated content.

For VLMA, responsible use of generative artificial intelligence should remain consistent with the principles underlying evidence-based Lifestyle Medicine: scientific integrity, professional competence, patient safety, transparency, and respect for people.

Generative artificial intelligence may help produce educational content more efficiently. It cannot take responsibility for whether that content is accurate, appropriate, or trustworthy.

That responsibility remains with the people and organizations choosing to publish it.

Selected References

1. International Committee of Medical Journal Editors. Recommendations for the Conduct, Reporting, Editing, and Publication of Scholarly Work in Medical Journals. Updated January 2026. Section V: Use of Artificial Intelligence in Publishing.

2. World Health Organization. Ethics and Governance of Artificial Intelligence for Health: Guidance on Large Multi-Modal Models. Geneva: World Health Organization; 2024. ISBN: 978-92-4-008475-9.

3. United Nations Educational, Scientific and Cultural Organization. Miao F, Holmes W. Guidance for Generative AI in Education and Research. Paris: UNESCO; 2023.

4. National Assembly of Viet Nam. Law on Artificial Intelligence No. 134/2025/QH15. December 10, 2025. Effective March 1, 2026.

5. Government of Viet Nam. Decree No. 142/2026/ND-CP Detailing Certain Articles and Measures for Implementing the Law on Artificial Intelligence. April 30, 2026. Effective May 1, 2026.

6. National Assembly of Viet Nam. Law on Personal Data Protection No. 91/2025/QH15. June 26, 2025. Effective January 1, 2026.

7. Government of Viet Nam. Decree No. 356/2025/ND-CP Detailing Certain Articles and Measures for Implementing the Law on Personal Data Protection. December 31, 2025. Effective January 1, 2026.

8. National Assembly of Viet Nam. Law on Intellectual Property No. 50/2005/QH11, as amended.

9. National Assembly of Viet Nam. Law No. 131/2025/QH15 Amending and Supplementing Certain Articles of the Law on Intellectual Property. December 10, 2025. Effective April 1, 2026.

10. Government of Viet Nam. Decree No. 134/2026/ND-CP Amending and Supplementing Certain Articles of Decree No. 17/2023/ND-CP on Copyright and Related Rights. April 6, 2026. Effective April 9, 2026.

11. National Assembly of Viet Nam. Law on Medical Examination and Treatment No. 15/2023/QH15. January 9, 2023. Effective January 1, 2024, as amended.

12. Rea BL, Cheema S, Lanza S, et al. Lifestyle Medicine Core Competencies: 2025 Update. American Journal of Lifestyle Medicine. 2026;20(3):443–451. First published online October 28, 2025. DOI: 10.1177/15598276251379821.

Editorial and Professional Scope Note

This article presents the professional perspective of the Vietnam Lifestyle Medicine Alliance on the responsible use of generative artificial intelligence in Lifestyle Medicine education as of October 2026. It is intended for professional information, education, and knowledge exchange.

It does not constitute individualized medical advice, a clinical artificial intelligence implementation guideline, a formal artificial intelligence risk assessment, an institutional compliance certification, or an authoritative legal interpretation. Its discussion of international guidance and Vietnamese legislation is educational and does not determine the legal classification or compliance status of any particular artificial intelligence system or activity.

VLMA is a physician-led professional platform operated by Vietnam Lifestyle Medicine Alliance Company Limited. It is not a medical examination or treatment service provider, artificial intelligence regulatory authority, medical technology certification body, or healthcare accreditation organization. VLMA does not independently diagnose patients, prescribe treatments, validate clinical artificial intelligence systems, or authorize their use in healthcare.

The use of generative artificial intelligence in education, research, publication, or healthcare must be assessed according to the actual purpose, parties involved, information processed, potential risks, and applicable Vietnamese laws and professional requirements. Where clinical decisions are involved, responsibility remains with appropriately authorized healthcare professionals and healthcare organizations.

References to international recommendations and governance frameworks are provided for educational purposes. They do not automatically create binding obligations under Vietnamese law or establish that VLMA has formally adopted, implemented, or certified the tools and practices discussed.